{"id":"W3174204476","doi":"10.18721/jhss.12209","title":"Searching for multicomponent terms in comparable scientific corpora","year":2021,"lang":"en","type":"article","venue":"St. Petersburg State Polytechnical University Journal","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Education and Early Childhood Development","funders":"","keywords":"Computer science; Linguistics; Epistemology; Natural language processing; History; Philosophy","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005054861,0.0006563087,0.0009949801,0.01164719,0.001868323,0.002792646,0.001417479,0.001001719,0.01409354],"category_scores_gemma":[0.02615204,0.0006481082,0.0008459972,0.01361913,0.0009776419,0.003449846,0.002837894,0.001310503,0.003599182],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001028691,"about_ca_system_score_gemma":0.002089787,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001430306,"about_ca_topic_score_gemma":0.002003477,"domain_scores_codex":[0.9950548,0.001449448,0.0008442652,0.001426732,0.001029131,0.0001955467],"domain_scores_gemma":[0.981611,0.01026295,0.001347334,0.003413374,0.003028123,0.0003372143],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00189526,0.0009612213,0.02525614,0.008873814,0.0007225221,0.01301452,0.01467496,0.006098987,0.1721485,0.05341034,0.02790066,0.675043],"study_design_scores_gemma":[0.0008684813,0.001287087,0.1294319,0.0018637,0.002198627,0.01868147,0.01585905,0.05689282,0.1614496,0.08743726,0.5236235,0.0004065132],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5647638,0.008389934,0.3305348,0.001803788,0.000802969,0.001983427,0.02632617,0.003376713,0.06201837],"genre_scores_gemma":[0.5813482,0.002911128,0.346288,0.0003282306,0.0004327666,0.001430903,0.05971637,0.001018115,0.006526381],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01409354,"threshold_uncertainty_score":0.04714763,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02146179342322107,"score_gpt":0.2643519645878402,"score_spread":0.2428901711646191,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}